Despite chemotherapy, patients with metastatic adenocarcinoma of the lung have a poor prognosis, therefore novel modalities such as immunotherapy are being developed for this disease. Two recent trials have been conducted utilizing GM-CSF-based tumor cell vaccines, with both immune and clinical responses reported. Colombo's group extended the concept of using GM-CSF by adding the expression of a potent activator of dendritic cells (DCs), CD40 ligand, in a murine model and showed that the combination was more effective than GM-CSF alone. Brenner's group showed that bystander cells transfected with the CD40 ligand gene admixed with tumor cells is an effective vaccine in a murine model. Based on these observations, we created a human bystander cell line that expresses both GM-CSF and CD40 ligand (GM.CD40L). This cell line enhanced the activation of an anti-tumor T cell response in an ex vivo human mixed autologous tumor cell/lymph node cell model. IL-12 and MJP-1a secretion, as well as expression of CD83, HLA-DR, and CD86 on DCs was induced confirming DC activation. We completed a phase I trial involving patients with solid tumors, testing GM.CD40L cells admixed with autologous tumor cells. The vaccine was safe, and the vaccine sites of these patients were densely infiltrated with activated (CD86 positive) DCs and T cells. We now propose to test the efficacy this vaccine in a phase II trial. The vaccine will consist of two components: (1) a mixture of two human lung adenocarcinoma cell lines that will serve as the source of tumor antigens, and (2) GM.CD40L cells to recruit and activate DCs at the vaccine site. Given the current view that vaccines given as a single agent have not and are unlikely to ever produce reasonable tumor response rates, we propose to incorporate all-trans retinoic acid (ATRA) to augment the effect of the vaccine. ATRA differentiates the immature myeloid cells (ImC) that are present in excess in cancer patients. ImC's accumulate as a result of tumor derived factor interference with DC differentiation, and these ImC in fact inhibit antigen specific T cell responses. Since our vaccine requires functional DCs, we propose to treat patients with ATRA before administering vaccines. In our pre-clinical model ATRA improved anti-tumor T cell responses induced by GM.CD40L cells. In addition to determining clinical efficacy, we will also determine if the strategy results in the activation of tumor antigen-specific T cell responses and an improvement in the DC: ImC ratio in treated patients. [unreadable] [unreadable] [unreadable]